LGTM.com - false positives: Useless comparison test

[kno10]

I see multiple false positives with this rule.

Apparently your parser fails to work correctly with ternary statements (I see many such false positives) or values that are tested multiple times in sequence. Yes, many people hat ternary expressions. But in hot code they can make a difference, and here maintenance is working fine. Its not as if boundary conditions in mathematics are going to change often.

https://lgtm.com/projects/g/elki-project/elki/snapshot/166e040796eaaf504872fe8869f4b45c0dc1776f/files/elki-core-math/src/main/java/elki/math/statistics/distribution/BetaDistribution.java#x43dfd0e3ba8863fe:1
https://lgtm.com/projects/g/elki-project/elki/snapshot/166e040796eaaf504872fe8869f4b45c0dc1776f/files/elki-core-math/src/main/java/elki/math/statistics/distribution/GammaDistribution.java#x8504d322f492f8ce:1
https://lgtm.com/projects/g/elki-project/elki/snapshot/166e040796eaaf504872fe8869f4b45c0dc1776f/files/elki-core-math/src/main/java/elki/math/statistics/distribution/GammaDistribution.java?sort=name&dir=ASC&mode=heatmap#xc1cdd586f50c276b:1
https://lgtm.com/projects/g/elki-project/elki/snapshot
/166e040796eaaf504872fe8869f4b45c0dc1776f/files/elki-core-math/src/main/java/elki/math/statistics/distribution/NormalDistribution.java?sort=name&dir=ASC&mode=heatmap#xa80c9b4c2e406404:1

This is different kind of false positive:
https://lgtm.com/projects/g/elki-project/elki/snapshot/166e040796eaaf504872fe8869f4b45c0dc1776f/files/elki-core-util/src/main/java/elki/utilities/datastructures/heap/Heap.java?sort=name&dir=ASC&mode=heatmap#x48a6e9c192ca5d50:1
which detects numeric overflow of int.

[aschackmull]

Thank you for your report. The first issue is not actually related to ternary expressions, but is rather a bug in bound strengthening in the range analysis related to inequalities. I've fixed both this and the FP for the overflow check in #2304 and #2305.